Fast content access and retrieval of JPEG compressed images

Autor: Farzad Zargari, Mohammad Ghanbari, Mahdi Mehrabi, Mohammad Amin Shayegan
Rok vydání: 2016
Předmět:
Zdroj: Signal Processing: Image Communication. 46:54-59
ISSN: 0923-5965
Popis: Fast content access and content based retrieval of images are among the common web and signal processing applications; on the other hand, JPEG is the dominant format for image compression in a wide range of applications. In this paper, a simple and fast method for content access and retrieval of JPEG coded images is presented which does not require complete decompression of coded images. The presented method uses DCT coefficients of coded blocks in the JPEG bit stream to extract average values of various size picture blocks namely DC values. The DC values provide an approximation of the coded image, which can be employed to construct a lower resolution picture or color histogram of the JPEG coded image for retrieval and other applications without full decompression of the image. A novel fast algorithm to access the content of JPEG compressed images is presented.Extracting 2×2 and 4×4 sub-block averages from JPEG bitstream without full decompression.The extracted sub-block averages can directly construct color histogram or an approximated picture.The presented method can be efficiently used for fast exploring or retrieval of JPEG coded images.Finding block averages from JPEG bitstream at the same accuracy as earlier works but faster.
Databáze: OpenAIRE